Molecular dynamics insights into CO₂/CH₄-n-docosane molecular mixing under deep reservoir conditions

Gas-oil miscibility within nanopores is a critical factor governing the efficient recovery of deep hydrocarbon resources. In this work, molecular dynamics (MD) simulations were performed to systematically investigate the miscibility, diffusion, and interfacial interaction mechanisms of carbon dioxide (CO₂) and methane (CH₄) with n-docosane (C₂₂H₄₆, abbreviated as C22) in hydroxylated silicon dioxide (SiO₂) nanopores (6-20 nm), explicitly accounting for the coupled effects of high temperature, high pressure, and pore-size confinement across a range of reservoir conditions (120 °C/30 MPa to 270 °C/80 MPa). Simulation results indicate that higher coupled temperature-pressure (T-P) states promote gas diffusion into heavy oil. However, this effect is counteracted by pore-wall adsorption and nanoconfinement, which suppress complete mixing. As the pore size decreases, the overlap of surface potential fields from opposing walls intensifies, inducing pronounced near-wall clustering and bridging configurations of C22 molecules. Mechanistically, CO₂ exhibits a superior capability to disrupt C22-C22 associations and weaken C22-SiO₂ adsorption, attributable to its stronger molecular affinity and polar interfacial interactions, which collectively facilitate the disaggregation and desorption of heavy components. In contrast, although CH₄ possesses a higher self-diffusion coefficient, its inherently weak intermolecular interactions render it unable to disrupt the ordered structure of heavy hydrocarbons, thereby limiting its ability to enhance C22 mobilization. This study elucidates the microscopic transport dynamics of heavy hydrocarbons under the synergistic control of nanopore confinement, thermal driving, and competitive gas-oil interactions, and crucially, identifies why CO₂ outperforms CH₄ in improving residual-oil mobility within deep nanopores. These findings provide molecular-scale insight into how gas type and nanoconfinement influence the redistribution and mobility of wall-associated long-chain hydrocarbons under reservoir-relevant conditions.